Legal claims defining the scope of protection, as filed with the USPTO.
1. A coding apparatus, comprising: a memory that stores instructions; a processor that executes the instructions; a first coder that encodes a low band part of an input audio signal to generate first encoded data which is transmitted to a decoder side, the low band part being equal to or lower than a predetermined frequency; and a second coder; wherein, when executed by the processor, the instructions cause the coding apparatus to perform operations comprising: normalizing a first spectrum included in each of sub-bands obtained from dividing the low band part, using a largest amplitude value among the first spectrum included in each sub-band, and obtaining a first normalized low-frequency spectrum, the first spectrum being obtained by decoding the first encoded data; calculating energy of the first spectrum in each of a plurality of the sub-bands obtained from dividing the low band part, smoothing the energy of the first spectrum in the sub-bands to obtain smoothened sub-band energy, and normalizing the first spectrum using the smoothened sub-band energy to generate a second normalized low-frequency spectrum; selecting a first normalizer that obtains the first normalized low-frequency spectrum or a second normalizer that generates the second normalized low-frequency spectrum based on a feature value of the first spectrum obtained by decoding the first encoded data; searching for a particular band where a value of correlation between a second spectrum and the first normalized low-frequency spectrum is highest among a plurality of candidate bands, the second spectrum being a spectrum in a high band part of the input audio signal that is higher than the predetermined frequency; and wherein the second coder encodes the second spectrum, and encodes information including an identifier of the particular band for transmitting the information to the decoder side, when the first normalizer is selected.
2. The coding apparatus according to claim 1 , wherein the second normalizer adds noise to the first spectrum.
3. The coding apparatus according to claim 1 , wherein the second normalizer performs clipping processing on the second normalized low-frequency spectrum.
4. A decoding apparatus, comprising: a memory that stores instructions; a processor that executes the instructions; a first decoder that receives an input first encoded data generated by encoding a low band part of an input audio signal in a coding apparatus and that decodes the first encoded data to generate a first spectrum, the low band part being equal to or lower than a predetermined frequency; and a second decoder; wherein, when executed by the processor, the instructions cause the decoding apparatus to perform operations comprising: normalizing a first spectrum included in each of sub-bands obtained from dividing the low band part, using a largest amplitude value among the first spectrum included in each sub-band, and obtaining a normalized low-frequency spectrum, the first spectrum being obtained by decoding the first encoded data; calculating energy of the first spectrum in each of a plurality of the sub-bands obtained from dividing the low band part, smoothing the energy of the first spectrum in the sub-bands to obtain smoothened sub-band energy, and normalizing the first spectrum using the smoothened sub-band energy to generate a normalized low-frequency spectrum; selecting a first normalizer that normalizes the first spectrum or a second normalizer that calculates energy of the first spectrum based on a feature value of the first spectrum obtained by decoding the first encoded data; wherein the second decoder receives as input the first normalized low-frequency spectrum or the second normalized low-frequency spectrum, and second encoded data generated in the coding apparatus, and decodes the second encoded data to generate a second spectrum, and wherein the second encoded data contains information indicating a particular band where a value of correlation is highest between an encoding-side second spectrum that is a spectrum in a high band part of the input audio signal that is higher than the predetermined frequency in the coding apparatus, and an encoding-side first spectrum that results from normalizing a spectrum generated by decoding the first encoded data in the coding apparatus.
5. A coding method, comprising: encoding, using a first encoder of a coding apparatus, a low band part of an input audio signal to generate first encoded data which is transmitted to a decoder side, the low band part being equal to or lower than a predetermined frequency; normalizing a first spectrum included in each of sub-bands obtained from dividing the low band part, using a largest amplitude value among the first spectrum included in each sub-band, and obtaining a first normalized low-frequency spectrum, the first spectrum being obtained by decoding the first encoded data; calculating energy of the first spectrum in each of a plurality of the sub-bands obtained from dividing the low band part, smoothing the energy of the first spectrum in the sub-bands to obtain smoothened sub-band energy, and normalizing the first spectrum using the smoothened sub-band energy to generate a second normalized low-frequency spectrum; selecting a first normalizer that obtains the first normalized low-frequency spectrum or a second normalizer that generates the second normalized low-frequency spectrum based on a feature value of the first spectrum obtained by decoding the first encoded data; searching for a particular band where a value of correlation between a second spectrum and the first normalized low-frequency spectrum is highest among a plurality of candidate bands, the second spectrum being a spectrum in a high band part of the input audio signal that is higher than the predetermined frequency; and encoding, using a second coder of the coding apparatus, the second spectrum, and encoding information including an identifier of the particular band for transmitting the information to the decoder side, when the first normalizer is selected.
6. A decoding method, comprising: receiving, at a first decoder of a decoding apparatus, an input first encoded data generated by encoding a low band part of an input audio signal in a coding apparatus, and decoding the first encoded data to generate a first spectrum, the low band part being equal to or lower than a predetermined frequency; normalizing a first spectrum included in each of sub-bands obtained from dividing the low band part, using a largest amplitude value among the first spectrum included in each sub-band, and obtaining a first normalized low-frequency spectrum, the first spectrum being obtained by decoding the first encoded data; calculating energy of the first spectrum in each of a plurality of the sub-bands obtained from dividing the low band part, smoothing the energy of the first spectrum in the sub-bands to obtain smoothened sub-band energy, and normalizing the first spectrum using the smoothened sub-band energy to generate a second normalized low-frequency spectrum; selecting a first normalizer that obtains the first normalized low-frequency spectrum or a second normalizer that generates the second normalized low-frequency spectrum based on a feature value of the first spectrum obtained by decoding the first encoded data; and receiving as input, at a second decoder of the decoding apparatus, the first normalized low-frequency spectrum or the second normalized low-frequency spectrum, and second encoded data generated in the coding apparatus, and decoding the second encoded data to generate a second spectrum, wherein the second encoded data contains information indicating a particular band where a value of correlation is highest between an encoding-side second spectrum, that is a spectrum in a high band part of the input audio signal that is higher than the predetermined frequency in the coding apparatus, and an encoding-side first spectrum that results from normalizing a spectrum obtained by decoding the first encoded data in the coding apparatus.
Unknown
July 5, 2016
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.